File Integrity Module

The file_integrity module sends events when a file is changed (created, updated, or deleted) on disk. The events contain file metadata and hashes.

The module is implemented for Linux, macOS (Darwin), and Windows.

How it works

This module uses features of the operating system to monitor file changes in realtime. When the module starts it creates a subscription with the OS to receive notifications of changes to the specified files or directories. Upon receiving notification of a change the module will read the file’s metadata and the compute a hash of the file’s contents.

At startup this module will perform an initial scan of the configured files and directories to generate baseline data for the monitored paths and detect changes since the last time it was run. It uses locally persisted data in order to only send events for new or modified files.

The operating system features that power this feature are as follows.

  • Linux - inotify is used, and therefore the kernel must have inotify support. Inotify was initially merged into the 2.6.13 Linux kernel.
  • macOS (Darwin) - Uses the FSEvents API, present since macOS 10.5. This API coalesces multiple changes to a file into a single event. Auditbeat translates this coalesced changes into a meaningful sequence of actions. However, in rare situations the reported events may have a different ordering than what actually happened.
  • Windows - ReadDirectoryChangesW is used.

The file integrity module should not be used to monitor paths on network file systems.

Configuration options

This module has some configuration options for tuning its behavior. The following example shows all configuration options with their default values for Linux.

- module: file_integrity
  paths:
  - /bin
  - /usr/bin
  - /sbin
  - /usr/sbin
  - /etc
  exclude_files:
  - '(?i)\.sw[nop]$'
  - '~$'
  - '/\.git($|/)'
  include_files: []
  scan_at_start: true
  scan_rate_per_sec: 50 MiB
  max_file_size: 100 MiB
  hash_types: [sha1]
  recursive: false
paths
A list of paths (directories or files) to watch. Globs are not supported. The specified paths should exist when the metricset is started.
exclude_files
A list of regular expressions used to filter out events for unwanted files. The expressions are matched against the full path of every file and directory. When used in conjunction with include_files, file paths need to match both include_files and not match exclude_files to be selected. By default, no files are excluded. See Regular expression support for a list of supported regexp patterns. It is recommended to wrap regular expressions in single quotation marks to avoid issues with YAML escaping rules.
include_files
A list of regular expressions used to specify which files to select. When configured, only files matching the pattern will be monitored. The expressions are matched against the full path of every file and directory. When used in conjunction with exclude_files, file paths need to match both include_files and not match exclude_files to be selected. By default, all files are selected. See Regular expression support for a list of supported regexp patterns. It is recommended to wrap regular expressions in single quotation marks to avoid issues with YAML escaping rules.
scan_at_start

A boolean value that controls if Auditbeat scans over the configured file paths at startup and send events for the files that have been modified since the last time Auditbeat was running. The default value is true.

This feature depends on data stored locally in path.data in order to determine if a file has changed. The first time Auditbeat runs it will send an event for each file it encounters.

scan_rate_per_sec
When scan_at_start is enabled this sets an average read rate defined in bytes per second for the initial scan. This throttles the amount of CPU and I/O that Auditbeat consumes at startup. The default value is "50 MiB". Setting the value to "0" disables throttling. For convenience units can be specified as a suffix to the value. The supported units are b (default), kib, kb, mib, mb, gib, gb, tib, tb, pib, pb, eib, and eb.
max_file_size
The maximum size of a file in bytes for which Auditbeat will compute hashes. Files larger than this size will not be hashed. The default value is 100 MiB. For convenience units can be specified as a suffix to the value. The supported units are b (default), kib, kb, mib, mb, gib, gb, tib, tb, pib, pb, eib, and eb.
hash_types
A list of hash types to compute when the file changes. The supported hash types are blake2b_256, blake2b_384, blake2b_512, md5, sha1, sha224, sha256, sha384, sha512, sha512_224, sha512_256, sha3_224, sha3_256, sha3_384, sha3_512, and xxh64. The default value is sha1.
recursive
By default, the watches set to the paths specified in paths are not recursive. This means that only changes to the contents of this directories are watched. If recursive is set to true, the file_integrity module will watch for changes on this directories and all their subdirectories.

Example configuration

The File Integrity module supports the common configuration options that are described under configuring Auditbeat. Here is an example configuration:

auditbeat.modules:
- module: file_integrity
  paths:
  - /bin
  - /usr/bin
  - /sbin
  - /usr/sbin
  - /etc